---
title: "DGEpi 2018 Dataquality Demo Dashboard"
output: 
  flexdashboard::flex_dashboard:
    orientation: columns
    vertical_layout: fill
    css: ["css/flex.css", "css/summarytools.css"]
    logo: pics/hzi_logo.png
    source_code: embed
    navbar:
      - { title: "Back to Talk", href: "index.html#23"}
---

```{r setup, include=FALSE}
library(flexdashboard)
library(plotly)
library(tidyverse)
library(summarytools)
library(formattable)
library(knitr)
library(kableExtra)
library(VIM)
library(lvplot)
df <- read_csv("data/outbreaks.csv")
```

Gesamt {data-navmenu="DQ Navigation" data-icon="fa-list" data-orientation=rows}
===

Column 
-----------------------------------------------------------------------

### Datenvollständigkeit

```{r}
DC <- 97
gauge(DC, min = 0, max = 100, symbol = '%', gaugeSectors(
  success = c(90, 100), warning = c(70, 89), danger = c(0, 69)
))
```

### Datengenauigkeit 

```{r}
DA <- 59
gauge(DA, min = 0, max = 100, symbol = '%', gaugeSectors(
  success = c(90, 100), warning = c(70, 89), danger = c(0, 69)
))
```

### gesamte Datenqualität

```{r}
DQ <- 85
gauge(DQ, min = 0, max = 100, symbol = '%', gaugeSectors(
  success = c(90, 100), warning = c(70, 89), danger = c(0, 69)
))
```

Column {data-width=350}
-----------------------------------------------------------------------

### Trend der Fälle in Prozent

```{r}
varis <- c("Florida", "California", "Ohio", "Illinois", "New York")
g2 <- df %>% 
  add_count(Year, State) %>% 
  add_count(State) %>% mutate(per = n/nn) %>% 
  select(Year, State, n, nn, per) %>% 
  filter(grepl(paste(varis, collapse = "|"), 
               State)) %>% 
  distinct() %>% 
  ggplot(aes(Year, per, color = State)) +
  geom_line() + theme_classic() + xlab("") +
  ylab("") + 
  scale_color_brewer(palette = "Set1", 
                     type ="qual") +
  
  scale_size_continuous(guide = "none")
ggplotly(g2) %>% 
  layout(legend = list(x = 0.75, y = 0.95))
```

### gesamte Fälle pro Jahr 

```{r}
df %>% 
  add_count(Year, State) %>% 
  add_count(State) %>%
  mutate(per = n/nn) %>% 
  select(Year, State, n, nn, per) %>% 
  distinct() %>% 
  group_by(Year) %>% 
  summarise(Cases_per_year = sum(n)) %>% 
  mutate(Cases_per_year = color_tile("white", "red")(Cases_per_year)) %>% 
  kable(escape = F, align = "c") %>%
  kable_styling(full_width = F)
```


Datenzusammenfassung {data-navmenu="DQ Navigation" data-icon="fa-list" data-orientation=columns}
===

Column {data-width=650}
-----------------------------------------------------------------------

### Chart A


```{r}
print(dfSummary(df, style = "grid", plain.ascii = FALSE, graph.magnif = 0.85), 
      method = "render", omit.headings = TRUE)
```



Datenvollständigkeit {data-navmenu="DQ Navigation" data-icon="fa-list" data-orientation=columns}
===

Column {data-width=650}
-----------------------------------------------------------------------

### Missing values per Variable (in Percent)

```{r}
g <- df %>% 
  summarise_all(funs(countNA)) %>% 
  gather(Variable, missings) %>% 
  mutate(miss_per = missings/nrow(df)*100) %>% 
  ggplot(aes(reorder(Variable, miss_per), miss_per, fill = miss_per)) +
  geom_bar(stat = "identity") +
  scale_fill_viridis_c(option = "C", guide = "none") +
  coord_flip() +
  theme_classic() +
  xlab("") +
  ylab("") +
  ggtitle("Missing values per Variable (in Percent)")
ggplotly(g) %>% 
  layout(showlegend = FALSE)
```

Column {data-width=350}
-----------------------------------------------------------------------

### Missing values per State (in Percent)

```{r}
g2 <- df %>% group_by(State) %>% 
  summarise_all(countNA) %>% 
  gather(Variable, missing, -State) %>% 
  group_by(State) %>% 
  summarise(miss = sum(missing)) %>% 
  left_join(df %>% add_count(State) %>% mutate(n = n*(ncol(df)-1)) %>% select(State, n) %>% distinct()) %>% 
  mutate(miss_per_s = round((miss/n*100), digits = 1)) %>% 
  filter(miss_per_s >= 31.6) %>% 
  ggplot(aes(reorder(State, miss_per_s), miss_per_s, fill = miss_per_s)) +
  geom_bar(stat = "identity") +
  
  scale_fill_viridis_c(option = "C", guide = "none") +
  
  coord_flip() +
  
  theme_classic() +
  xlab("") +
  ylab("") +
  ggtitle("Missing values per State (in Percent)")
ggplotly(g2) %>% 
  layout(showlegend = FALSE)
```

### Gesamt Datenvollständigkeit

```{r}
DC <- 97
valueBox(DC, icon = "fa-thumbs-up", color = "green")
```

Datengenauigkeit {data-navmenu="DQ Navigation" data-icon="fa-list" data-orientation=columns}
===

Column {data-width=650}
-----------------------------------------------------------------------

### Anzahl der Erkrankten pro Ausbruch nach US Staat

```{r}
df %>% 
  select(State, Illnesses) %>% 
  add_count(State) %>% 
  filter(n >= 1000) %>% 
  ggplot(aes(State, Illnesses)) +
  geom_lv( outlier.colour = "red", fill = "#005aa0", color = "#005aa0") +
  theme_classic() +
  xlab("") +
  ylab("")
```

Column {data-width=350}
-----------------------------------------------------------------------

### Chart B

```{r}
test <- df %>% select(Year, Ingredient, Species, Status) %>% as.matrix()
pbox(test)
```

### Chart C

```{r}
DC <- 59
valueBox(DC, icon = "fa-thumbs-down", color = "red")
```